Composite True Position Calculations

[PCDMISRON]

I am using v4.1+ on a job and I am not confident that PC_DMIS is accurately reporting the composite true position in the program in which I am applying the tolerancing. Could someone please describe to me how I could manually calculate this measurement in order to verify the correctness of teh results I am getting from PC_DMIS?

Thanks,
Ron

[Jan d.]

Can you give us the exact call out? Also, please provide the EXACT call for the datums (do they have VC's????????)!

I have done composites a lot and they seem to work for me.

Be careful though with datums that have multiple VCs. There is a bug that will be fixed in V4.2.


Jan.

[JoeRZ]

I don't use 4.1 but per ANSI standard -

Composite Positional Tolerance is used to locate the entire feature pattern as well as define the position and orientation of each of the features in the pattern set. The top part of the callout governs the individual holes, while the bottom part is applied to the entire pattern.

First create a true pos dimension for each of the features selecting the appropriate datum frame i.e., A,B,C, or from features created that represent 3-2-1 alignment.

The second part is for the position of the features relative to each other as opposed to the datum frame. Be sure features have nominal data - necessary for best-fit alignment.
Insert a 2-D best-fit alignment using features measured in pattern - make sure that 'Translate and Rotate' are selected.
Dimension the position of each feature in the pattern. Be sure to un-check 'Use Datums' when creating dimension.

The dimension will report the position of the circles relative to each other via the 2-D Best Fit Alignment. Once the position is completely dimensioned be sure to recall alignment before dimensioning other features in the program.

The formula for TP is 2 * square root of X deviation squared + Y deviation squared.

Hope this helps…

JRZ

[PCDMISRON]

Can you give us the exact call out? Also, please provide the EXACT call for the datums (do they have VC's????????)!

I have done composites a lot and they seem to work for me.

Be careful though with datums that have multiple VCs. There is a bug that will be fixed in V4.2.


Jan.

Hi Jan, thanks for your interest. The callouts are an example of what I would like to verify. The upper callout (PLTZF) is a tolerance of 0.016 @ MMC Datum Ref. A-B-C and the lower callout (FRTZF) tolerance of 0.006 @MMC Datum Ref. A-B. I know how to calculate actual TP but I a not certain how to calculate the hole-to-hole over a three-hole pattern.

The data below is taken from a spreadsheet that I put together in an attempt to list the specifics of this case. My apologies for the appearance of the format of the information.

If you can picture this example, the information for which I am needing to know the calculations is last column on the second you of data, the hole-to-hole actual TP. If more info is needed, please let me know. Again, thanks for your help.

Hole No. Hole @ MMC Hole Actual Size Pattern Tol Allowed
1 0.310 0.311 0.017
2 0.310 0.312 0.018
3 0.310 0.315 0.021

Actual T.P. Hole-To-Hole Tol Hole-To-Hole Actual T.P.
0.008 0.007 0.007
0.017 0.008 0.008
0.021 0.011 0.011

Distance To Datum B Actual Distance To Datum B
1.000 1.004
2.000 2.006
3.000 3.010

Distance To Datum C Actual Distance To Datum C
1.000 1.000
7.000 7.006
1.000 0.999

BTW, what are VC's?

[JoeRZ]

VC = Virtual Condition that takes the Geo Tol and feature modifiers into account when calculating for fit / function. What Jan was saying about multiple VC is if datums are features of size, need to incorporate in formula.

JRZ

[PCDMISRON]

VC = Virtual Condition that takes the Geo Tol and feature modifiers into account when calculating for fit / function. What Jan was saying about multiple VC is if datums are features of size, need to incorporate in formula.

JRZ

Tahnks, again Joe...Having a real brainfart of a day...must be the holidays...But can you give me the components as applied in order to calculate the measured TP for the FRTZF? I am feeling really slow about how to verify this number.

Ron

[tnguyen]

Hi JoeRZ . I have the same situation with Ron, but the lower callout only Datum A. Could you explain more details how to apply for this ?.
Thanks

[Jan d.]

OK, here's my attempt at an explanation; probably not very scientific, but here we go.

I am going to presume that A is a plane and B and C are 2 holes (features of size). In your call-out it appears as if none of the datums are evlauated with a virtual condition. That makes it a lot easier.

The way I do this is that I treat both of them independently. I evaluate the feature to ABC and then I evaluate one more time to A and B. Again, V4.1 makes this a super simple breeze.

Now the way you ought to envision that, in my mind, is a hard gauge. ABC set up a hard gauge. Envision a flat plate with 2 pins coming out where B and C are (size at RFS; or an expanding pin in other words; this sets up 1 unique orientation for you hard gauge, based on the actual measured datums). The pattern holes are at MMC where the holes are supposed to be based on the basic dimensions. After having the hard gauge engage the datums on the part, you try to stick in the pins (size at hole MMC). If they slide freely though, you are OK. If they do not, you fail the TP test.

Now depending on the geometrical lay-out, the ultimate TP band around one hole might be different from one hole to the other. Envision B and C on one side of the part, while you holes are on the other. If B and C set up an angular error, the positional error becomes bigger as you move away from your datum structure. So the hole furthest away from the datum frame would have the closest requirements. Only if B was at the center if your pattern, would the TP on all 3 holes be the same.

Again, I tried this with V4.1, and it does calculate different numbers for every single hole in the pattern. It looks as if it is also in the correct direction and by the correct amount. I did not mathematically verify this.

If you have a lot of these, you ought to upgrade to 4.1 because it will make your life much easier.



Jan.

[JoeRZ]

For 1 Datum Feature

The lower segment specifies hole to hole requirement and orientation of the requirement to the datums.

The FRTZF may skew, rotate and/or be displaced within the confines of the PLTZF. The axes of the holes must lie in both zones simultaneously.

Meaning you can wiggle the part to fit within PLTZF but must meet FRTZF.

For FRTZF, you need to compare hole to hole distance requirement (evaluate the distance between holes compared to nominals), but will need to be best fit. You will need to determine angle of displacement first to reduce amount of deviation for TP calculation.

For instance (take note that this is an approximation because best fit not included):

Hole 1 to hole 2 from datum B difference = 0.002
Hole 1 to hole 2 from datum C difference = 0.006

Hole 1 to Hole 2 TP = 0.012 (most likely value is reduced if best fit) and so on…

For 2 Datum Feature

The FRTZF may move up and down or left and right within the confines of the PLTZ but it may not skew or rotate. The axes of the holes may only rotate within the confines of the FRTZF.

For Virtual Condition-

If checking hole - Pin would be MMC of hole minus Geometric tolerance.
If checking pin - Ring would be MMC of pin plus Geometric tolerance.

Works good with gage design - functional gage.

Hope this helps...

JRZ

[tnguyen]

Thanks

[PCDMISRON]

Thaks, Jan!!

Jan, I want to thank you for the information that you provided in helping me to understand the actual calculations for the FRTZF protion of the composite True Postion callout from the measured dimensions. Your knowledge went a long way in conjunction with that of others to helping me learn how to verify the output from v4.1
Thanks, again
Ron

OK, here's my attempt at an explanation; probably not very scientific, but here we go.

I am going to presume that A is a plane and B and C are 2 holes (features of size). In your call-out it appears as if none of the datums are evlauated with a virtual condition. That makes it a lot easier.

The way I do this is that I treat both of them independently. I evaluate the feature to ABC and then I evaluate one more time to A and B. Again, V4.1 makes this a super simple breeze.

Now the way you ought to envision that, in my mind, is a hard gauge. ABC set up a hard gauge. Envision a flat plate with 2 pins coming out where B and C are (size at RFS; or an expanding pin in other words; this sets up 1 unique orientation for you hard gauge, based on the actual measured datums). The pattern holes are at MMC where the holes are supposed to be based on the basic dimensions. After having the hard gauge engage the datums on the part, you try to stick in the pins (size at hole MMC). If they slide freely though, you are OK. If they do not, you fail the TP test.

Now depending on the geometrical lay-out, the ultimate TP band around one hole might be different from one hole to the other. Envision B and C on one side of the part, while you holes are on the other. If B and C set up an angular error, the positional error becomes bigger as you move away from your datum structure. So the hole furthest away from the datum frame would have the closest requirements. Only if B was at the center if your pattern, would the TP on all 3 holes be the same.

Again, I tried this with V4.1, and it does calculate different numbers for every single hole in the pattern. It looks as if it is also in the correct direction and by the correct amount. I did not mathematically verify this.

If you have a lot of these, you ought to upgrade to 4.1 because it will make your life much easier.



Jan.

[PCDMISRON]

Thanks, Joe

Thanks, Joe...The info helped a ton.

Ron

For 1 Datum Feature

The lower segment specifies hole to hole requirement and orientation of the requirement to the datums.

The FRTZF may skew, rotate and/or be displaced within the confines of the PLTZF. The axes of the holes must lie in both zones simultaneously.

Meaning you can wiggle the part to fit within PLTZF but must meet FRTZF.

For FRTZF, you need to compare hole to hole distance requirement (evaluate the distance between holes compared to nominals), but will need to be best fit. You will need to determine angle of displacement first to reduce amount of deviation for TP calculation.

For instance (take note that this is an approximation because best fit not included):

Hole 1 to hole 2 from datum B difference = 0.002
Hole 1 to hole 2 from datum C difference = 0.006

Hole 1 to Hole 2 TP = 0.012 (most likely value is reduced if best fit) and so on…

For 2 Datum Feature

The FRTZF may move up and down or left and right within the confines of the PLTZ but it may not skew or rotate. The axes of the holes may only rotate within the confines of the FRTZF.

For Virtual Condition-

If checking hole - Pin would be MMC of hole minus Geometric tolerance.
If checking pin - Ring would be MMC of pin plus Geometric tolerance.

Works good with gage design - functional gage.

Hope this helps...

JRZ

[d.evans]

The second part is for the position of the features relative to each other as opposed to the datum frame. Be sure features have nominal data - necessary for best-fit alignment.
Insert a 2-D best-fit alignment using features measured in pattern - make sure that 'Translate and Rotate' are selected.
Dimension the position of each feature in the pattern. Be sure to un-check 'Use Datums' when creating dimension.

The dimension will report the position of the circles relative to each other via the 2-D Best Fit Alignment. Once the position is completely dimensioned be sure to recall alignment before dimensioning other features in the program.

JRZ

How would you calculate the TP of the second part without a CMM?

for example, you have a rectangular 1" thick plate with a hole pattern, sort of like the old B&S cheese block. How would you figure the hole pattern 'to itself' TP using just a surface plate and hand tools?

This one has had me stumped for a while.

[T Miller]

Graft paper and onion skin paper. You plot the second call out on the onion skin and lay over the plot on the graft for the top call out. Draw your diameteters to scale to call out then plot your points where they lie - you can move the onion skin plot around as long as it falls within the graft

[craiger_ny]

Hard gage, it is what GD&T is currently built on. Yes it has been and will continue to make provisions for soft gages but the foundation of GD&T is a hard gage. That is how you'd do it and anything else for that matter. Coventional direct reading gages are very tricky for meeting the true interpretation of GD&T.

Craig